Humanity announced on Wednesday that it will launch a new, highly expensive subscription plan for AI chatbot Claude: Max. Max, an answer to Openai’s 200-month ChatGpt Pro tier, has higher usage restrictions than humanity’s $20/month Claude Pro subscription, and prioritized access to the company’s latest AI models and features.
A bit confusing, Max has two flavors with different prices and usage restrictions. There’s a maximum tier of $100 per month with a rate limit five times higher than the Claude Pro, and a monthly maximum option with a rate limit 20 times higher.
Frontier AI model developers are looking for new ways to increase revenue. Also, offering expensive AI subscriptions to power users seems like a promising way to do that. Just two months after launching the ChatGpt Pro, Openai reportedly told investors that it had increased its annual revenue by $300 million. If humanity’s new maximum plan is similarly successful, it could give the company a much-needed boost.
Anthropic product lead Scott White told TechCrunch in an interview that the company has not one day set up more expensive subscriptions to rule out. In particular, Anthropic doesn’t have any unlimited usage plans yet. This is provided by Openai, which is offered with ChatGpt Pro.
When asked if humanity would consider launching a $500/month Claude plan, White said, “We always have many exploratory options available.” User feedback continues to guide the human product roadmap, he added.
To meet the enormous costs of developing frontier AI models, humanity is investigating a variety of new revenue channels. One of them is Claude for education. Claude for Education targets university clients and offers the university specific capabilities and advantages.
Humanity refused to tell TechCrunch how many Claude subscriptions they have sold so far. However, White said the company’s new Claude 3.7 Sonnet AI model created “a lot of demand” for its products. Claude 3.7 Sonnet is the first human reasoning model, the type that uses computing power to reliably answer specific questions than traditional models.
Source link
